Optimization Design of Dual-Redundancy Permanent Magnet Synchronous Motor Based on Improved Iterations Taguchi Method

Compared with the permanent magnet synchronous motor (PMSM), the dual redundancy permanent magnet synchronous motor (DRPMSM) has more working states and optimized design parameters. Most of the multi-objective optimization methods for DRPMSM are based on the optimization design methods of PMSM, which is difficult to meet the optimization requirements of DRPMSM in multiple working states. Moreover, the traditional Taguchi method cannot optimize the gap between the two-level values of the parameters. Therefore, this paper proposes an improved iteration Taguchi method for the DRPMSM, combining the Taguchi method with a sequential optimization strategy. The sensitivity of parameters based on the Pearson correlation coefficient is filtered. The effectiveness of the proposed method is verified by the finite element analysis (FEA) and experimental results. Firstly, the sensitivity of parameters was calculated, and the parameters were screened for the single redundancy working state and dual redundancy working state, respectively. Secondly, the Taguchi method was used in the single redundancy working state. Four groups of parameters were obtained, namely, the optimal efficiency, the optimal torque, the optimal torque ripple coefficient, and the optimal Taguchi. Next, the gap of two-level values of the parameters was scaled down under the dual redundancy working state. In addition, the Taguchi method was used again to optimize the DRPMSM motor in the dual redundancy working state. Finally, the optimal values of the parameters were selected. In the improved iterative Taguchi method, the parameters set obtained from single redundancy working state limits the gap of two-level values of the dual redundancy working state parameters. The issue that the traditional Taguchi method cannot optimize the gap of two-level values was addressed. Simulation results on the DRPMSM show that, for single redundancy working state, the efficiency of the motor is increased by 0.75 %, the output torque is increased by 1.16 %, and the torque ripple coefficient is decreased by 27.42 %. Furthermore, for dual redundancy working state, the efficiency of the motor is increased by 1.86 %, the output torque is increased by 0.62 %, and the torque ripple coefficient is decreased by 24.07 %. The error of the FEA results is less than 10 % compared with the experimental test of the prototype. The error of efficiency and the error output torque of the FEA and experimental are less than 2 %. The error of the FEA and experimental torque ripple coefficient is large due to the axial deviation and coupler deformation. The design motor performances in single and dual redundancy working states are close to that of the experimental test of the prototype, which indicates that the prototype can maintain the corresponding performance in single and dual redundancies. The following conclusions can be drawn from the simulation analysis and the prototype experiment. (1) The introduction of the Pearson correlation coefficient is helpful to select parameters. (2) The combination of the Taguchi method and sequential optimization strategy can address the issue of optimizing the gap of two-level values of the parameter in the Taguchi method. Different working states of the motor can also be considered. (3) The parameters with high sensitivity under different states can be considered, and the selection among parameters is comprehensive. © 2023 Chinese Machine Press. All rights reserved.